Directed, heritable changes in gene expression, known as epigenetic changes, are thought to account for the stability of the differentiated state. Our working hypothesis is that tumor transformation in higher plants, and perhaps generally, is due to epigenetic changes rather than from permanent changes in the cell genome. We are using plant cells in culture to find molecular mechanisms for epigenetic changes that perpetuate the tumor state in crown-gall, a neoplastic disease of plants. Our immediate objective is to test the hypothesis that tumor autonomy is maintained in dividing populations of cells by a positive feedback loop in which cell division factors induce their own production by the cells.